ISSN 0023-1584, Kinetics and Catalysis, 2020, Vol. 61, No. 3, pp. 374–383. © Pleiades Publishing, Ltd., 2020.

Russian Text © The Author(s), 2020, published in Kinetika i Kataliz, 2020, Vol. 61, No. 3, pp. 349–358.

A Mathematical Model of the Synthesis of Pentaerythritol

D. I. Belkina, * and O. A. Demchenkoa
a
Institute of Chemical Technologies, Dahl East Ukrainian National University, Rubezhnoye, 93009 Ukraine
*e-mail: dib@rambler.ua
Received October 11, 2019; revised December 2, 2019; accepted December 13, 2019

Abstract—Based on the results of a kinetic study of the condensation of acetaldehyde and formaldehyde in
the presence of sodium hydroxide and taking into account well-known data, we compiled a reaction scheme
of the process, in which pentaerythritol, dipentaerythritol, tripentaerythritol, bispentaerythritolformal, pen-
taerythritol methyl ether, and sodium 2,2-bis(hydroxymethyl)propanoate are formed. According to the reac-
tion scheme, we developed a mathematical model of the process, which includes reaction rate equations and
formulas for calculating the activity coefficients of ions and the degrees of ionization and equilibrium con-
centrations of intermediate substances. The effective kinetic parameters and activation energies of reactions
were found. The adequacy of the mathematical model was statistically evaluated under the following condi-
tions: an acetaldehyde–formaldehyde–sodium hydroxide molar ratio of 1 : 4–9 : 1.2, an initial acetaldehyde
concentration of 0.4–0.9 mol/L, and a temperature of 13–47°C.

Keywords: pentaerythritol, synthesis, mathematical model, kinetic parameters

DOI: 10.1134/S0023158420030052

INTRODUCTION —a sequential reaction of the aldol condensation of

Pentaerythritol (2,2-bis(hydroxymethyl)-1,3-pro- acetaldehyde (A) and formaldehyde (F) with the forma-
panediol, P), which is used for the manufacture of var- tion of 3-hydroxypropanal (A1), 2-hydroxymethyl-3-
nishes, resins, polymers, and other products, is syn- hydroxypropanal (A2), and 2,2-bis(hydroxymethyl)-
thesized by the condensation of acetaldehyde and 3-hydroxypropanal (A3) [3];
formaldehyde in the presence of sodium hydroxide. —the cross Cannizzaro reaction of 2,2-
Due to the high reactivity of starting and intermediate bis(hydroxymethyl)-3-hydroxypropanal and formal-
aldehydes, the process is a complex reaction system, dehyde with the formation of pentaerythritol and
and a study of its kinetics is associated with the use of sodium formate (FN) [3];
a mathematical model. We failed to find data on such —side reactions with the formation of dipen-
studies in the literature. taerythritol (2,2,6,6-tetrakis(hydroxymethyl)-4-oxa-
Earlier [1, 2], we conducted a study of this kind 1,7-heptanediol) (DP) and bispentaerythritolformal
with the development of a mathematical model for the (2,2,8,8-tetrakis(hydroxymethyl)-4,6-dioxanonane-
synthesis of pentaerythritol, which took into account diol) (BP), the starting materials of which are 2-pro-
the following reactions: penal (A1X), 2-hydroxymethyl-2-propenal (A2X),
Symbols and abbreviations: A, acetaldehyde; AA, 3-hydroxybutanal; A1, 3-hydroxypropanal; A2, 2-hydroxymethyl-3-hydroxypro-
panal; A3, 2,2-bis(hydroxymethyl)-3-hydroxypropanal; AD, 2,4-dimethyl-6-hydroxy-1,3-dioxacyclohexane; AD1, 2-(2-hydroxypro-
pyl)-4-methyl-6-hydroxy-1,3-dioxacyclohexane; A1X, 2-propenal; A2X, 2-hydroxymethyl-2-propenal; D2, 2-hydroxymethyl-4-oxa-
6,6-bis(hydroxymethyl)-7-hydroxyheptanal; B2, 2-hydroxymethyl-4,6-dioxa-8,8-bis(hydroxymethyl)-9-hydroxynonanal; M2, 2-
hydroxymethyl-4-oxapentanal; T2, 2-hydroxymethyl-4,8-dioxa-6,6,10,10-tetrakis(hydroxymethyl)-11-hydroxyundecanal; AE, A1E,
A2E, D2E, B2E, M2E, and T2E, corresponding enolate ions; D1E, the enolate ion of 4-oxa-6,6-bis(hydroxymethyl)-7-hydroxyhep-
tanal; B1E, the enolate ion of 4,6-dioxa-8,8-bis(hydroxymethyl)-9-hydroxynonanal; M1E, the enolate ion of 4-oxapentanal; T1E, the eno-
late ion of 4,8-dioxa-6,6,10,10-tetrakis(hydroxymethyl)-11-hydroxyundecanal; D3, 2,2,6,6-tetrakis(hydroxymethyl)-4-oxa-7-hydroxyhep-
tanal; B3, 2,2,8,8-tetrakis(hydroxymethyl)-4,6-dioxa-9-hydroxynonanal; M3, 2,2-bis(hydroxymethyl)-4-oxapentanal; T3, 2,2,6,6,10,10-
hexakis(hydroxymethyl)-4,8-dioxa-11-hydroxyundecanal; P, pentaerythritol; PN, sodium 2,2-bis(hydroxymethyl)propanoate; FN, sodium
formate; BP, bispentaerythritolformal; F, formaldehyde; B, sodium hydroxide; M, methanol; DP, dipentaerythritol; TP, tripentaerythri-
tol; MP, pentaerythritol methyl ether; PF, pentaerythritol formaldehyde hemiacetal; FM, methanol formaldehyde hemiacetal; TF,
formaldehyde trimer; AL, aldehydes; M, formaldehyde to acetaldehyde molar ratio; KA, equilibrium constant of aldol condensation;
k, rate constant; Z, ratio between the rate constants of fast reactions; S, complex associated with the steady-state concentration of eno-
late ions; pKa(i), ionization constant; α, degree of ionization; γ, ionic activity coefficient; z, ion charge; f, number of degrees of free-
dom; S2, dispersion; N, number of experiments; n, number of factors; FT, Fisher test value; e, experimental; c, calculated; O, reverse reac-
tion; (0), initial concentration; s, total concentration; (A), AL consumption in reactions not accompanied by the consumption of B.

374
 A MATHEMATICAL MODEL OF THE SYNTHESIS OF PENTAERYTHRITOL
Table 1. Experimental conditions
Levels of factors M = СF(0)/CA(0) CA(0), mol/L T, °C
0 6.5 0.65
+ 8.0 0.80
– 5.0 0.50
+1.68 9.0 0.90
–1.68 4.0 0.40
and ionized pentaerythritol and its formaldehyde
hemiacetal (PF) [2, 4, 5];
—side reactions of acetaldehyde with the formation
of cyclic hemiacetals, in particular, aldoxane (2,4-
dimethyl-6-hydroxy-1,3-dioxacyclohexane) (AD1) [6].
However, the inadequacy of the mathematical
model under certain process conditions (without
adjusting it with empirical coefficients) showed that it
does not take into account some features of the pro-
cess that are not predicted by the set of known data.
In this regard, we conducted a kinetic study of the
process in order to identify the causes of these features,
to make necessary changes in the reaction scheme,
and to develop an adequate mathematical model of the
test process.
EXPERIMENTAL
In the experiments, we used liquid acetaldehyde,
formalin with formaldehyde and methanol weight
fractions of 23–37 and 1–10%, respectively, and an
aqueous solution with a sodium hydroxide weight
fraction of 40%. The condensation process was carried
out in a thermostatically controlled glass flask
equipped with a stirrer. The calculated amounts of the
aqueous solution of sodium hydroxide and formalin
were introduced into the flask. Then, acetaldehyde
was rapidly added with vigorous stirring.
The experiments were carried out at an acetalde-
hyde–formaldehyde–sodium hydroxide molar ratio
of 1 : 3–10 : 0.8–1.4), an initial acetaldehyde concen-
tration of 1.2–0.26 mol/L, and a temperature of 5–
50°C.
KINETICS AND CATALYSIS Vol. 61 No. 3 2020
375
In the course of the process, the reaction mixture
was sampled, and the concentrations of identified
products in the samples were determined by gas–liq-
uid chromatography [7]. To determine the concentra-
tions of sodium hydroxide and aldehydes, a sample of
30 the reaction mixture was transferred to a flask with a
specified amount of hydrochloric acid solution; a phe-
40 nolphthalein indicator was added, and the acid was
titrated with a 1 M solution of sodium hydroxide until
a slight pink color appeared. Then, a solution of
20 hydroxylamine hydrochloride and a bromophenol
blue indicator were introduced into the flask. After a
47
30-min exposure, the sample was titrated with a 1 M
solution of sodium hydroxide until a slight bluish color
appeared.
13
To evaluate the adequacy of the mathematical
model, we carried out the experiments according to a
compositional plan under the conditions given in
Table 1 at a constant molar ratio of 1 : 1.2 between
acetaldehyde and sodium hydroxide to complete the
process. (An increase in this ratio had almost no effect
on the yield of products under the conditions of a
batch process.) In these experiments, formalin with 37
and 7% weight mass fractions of formaldehyde and
methanol, respectively, was used.
RESULTS AND DISCUSSION
Below is given a generalized reaction scheme of the
process of acetaldehyde and formaldehyde condensa-
tion in the presence of sodium hydroxide. It was com-
piled based on well-known data and our experimental
results with consideration for a preliminary assessment
of the adequacy of the developed mathematical model
under different process conditions. Based on these
results, Scheme 1 was supplemented by a dispropor-
tionation reaction of 2,2-bis(hydroxymethyl)-3-
hydroxypropanal, in which pentaerythritol and
sodium 2,2-bis(hydroxymethyl)-3-hydroxypropa-
noate (PN) (which was not identified previously as a
by-product of the process) are formed, reactions of the
formation of tripentaerythritol and pentaerythritol
methyl ether, and reactions of formaldehyde with the
production of its trimer and hemiacetals, which deter-
mine an equilibrium concentration of this aldehyde.
For brevity, Scheme 1 shows the symbols of substances
and rate constants.
376 BELKIN, DEMCHENKO

CH3CHO H3C OH
(AD)
(A) O O
(A), B
k1, B k1O
k11 kAD CH3 H3C OH
(HOCH2)4C (P)
CH3CHO CH(OH)CH3 (AA), B (AD1) (HOCH2)3CCOONa
−
H2C CHO O O
B kAD1 (PN)
(AE) CH2CHO
k11O
(AA) H3C(OH)HCCH2 (R)3CCHO, k42
k1E, F B, k1EO
NaOH
k3E, F
RCH2CHO k 2, B RHC−CHO k2E, F
(R)2HCCHO
k3, B
(R)2C−CHO (R)3CCHO
(A1) k2O (A1E) k2EO, B (A2) k3O
(A2E) k3EO, B (A3)
−
k1X OH R: −CH2OH k2X OH− NaOH k41
k1XO F
R1: −H k1XO
HCCHO R2: −C(CH2OH)3 (HOCH2)4C
R
CH2 (P)
R3: −OCH2C(CH2OH)3 H2C CCHO
(A1X) HCOONa
R4: −C(R)2CH2OCH2C(R)3 (A2X)
R1−4CH2O−

kD1, kD1O,
kB1,

R1−4CH2O−
kB1O, kD2, kD2O,
kM1, kM1O, kD1E, k2D, kB2, kB2O,
kT1 kT1O kB1E, k2B, kM2, kM2O,
kM1E, R k2M, kT2 kT2O
kT1E k2T
−
H2CHC CHO F H2CHCCHO B
R
OCH2R1−4 B
kD1EO,
OCH2R1−4 kDO,
kBO,
H2CC−HO
(D1E), (B1E), kB1EO, (D2), (B2), kMO,
kM1EO, (M2), (T2) kTO OCH2R1−4
(M1E), (T1E)
kT1EO
(D2E), (B2E),
(M2E), (T2E)
kD2E, kD2EO,
kB2E, kB2EO,
F B
kM2E, kM2EO,
HCHO + CH2(OH)2 kT2E kT2EO
(F)
NaOH

NaOH
kF C(R)3 (B) C(R)2CHO
OCH2R1−4 kD, kB, kM, kT OCH2R1−4
HCOONa + CH3OH
HCOONa (D3), (B3),
(DP), (BP), (M3), (T3)
(MP), (TP)

2CH2(OH)2 HOCH2OCH2OH + CH2(OH)2 HOCH2OCH2OCH2OH

H 2O H 2O
(TF)
OCH3 nF, KPF
HCHO + CH3OH H2C (HOCH2)4C (HOCH2)4 −nC(CH2OCH2OH)n
OH
(PF)
(FM)
Scheme 1. Reaction scheme of the test process with the symbols of substances and reaction rate constants.

Scheme 1 corresponds to the system of reaction k2C A1 + k2EOC A2 + k1XOC A1X COH− CB
rate equations (1)–(37) given below. The concentra- S2 =
tions of enolate ions involved in the fast stages of the 1 + Z2CF + Z4 (2)
aldol condensation reactions are the following steady- = k2OC A1E CB ,
state concentrations (stationary concentration
method): k3C A2 + k3EO + k2XOC A2X COH− CB
S3 =
k1C A + k11O + k1EO 1 − Z3CF + Z5 (3)
S1 = = k1OC AE CB , (1)
1 − ZC A + Z1CF = k3OC A2E CB ,

KINETICS AND CATALYSIS Vol. 61 No. 3 2020

 A MATHEMATICAL MODEL OF THE SYNTHESIS OF PENTAERYTHRITOL 377

kD1CPC A1X (COH− CB ) + kD1EOCD2 dC AD1 dt = kAD1C AA

2
CBγ, (18)
S4 =
1 − Z6CF (4)
dC A1X dt = (Z 4S2 + S4 + S6 + S8 + S10 )CBγ
= kD1OCD1E CB ,
− (k1XOC A1X + kD1CPC A1X + kB1CPFC A1X (19)
kD2CPC A2X (COH− CB )
S5 = = kD2OCD2E CB , (5) + kM1CMC A1X − kT1CDPC A1X )COH− γ,
1 − Z7CF + Z8
dC A2X dt = (Z 5S3 + S5 + S7 + S9 + S11)CBγ
kB1CPFC A1X (COH− CB ) + kB1EOCB2
S6 = − (k2XOC A2X + kD2CPC A2X + kB2CPFC A2X (20)
1 − Z9CF (6)
+ kM2CMC A2X + kT2CDPC A2X )COH− γ,
= kB1OCB1E CB ,
kB2CPFC A2X (COH− CB ) + kB2EOCB3 + k2BCB2 dCD2 dt = (Z 6S4CF − kD1EOCD2
S7 = − k2DCD2 + Z 8S5)CBγ,
(21)
1 − Z10CF + Z11 (7)
= kB2OCB2E CB , dCB2 dt = (Z 9S6CF − kB1EOCB2
(22)
kM1CMC A1X (COH− CB ) + kM1EOCM2 − k2BCB2 + Z11S7 )CBγ,
S8 =
1 − Z12CF (8)
dCM2 dt = (Z12S8CF − kM1EOCM2
= kM1OCM1E CB , (23)
− k2MCM2 + Z14S9 )CBγ,
kM2CMC A2X (COH− CB ) + kM2EOCB3 + k2MCM2
S9 = dCT2 dt = (Z15S10CF − kT1EOCT2
1 + Z13CF + Z14 (9) (24)
− k2TCT2 + Z17S11)CBγ,
= kM2OCM2E CB ,
kT1CDC A1X (COH− CB ) + kT1EOCT2 dCD3 dt = (Z 7S5CF − kD2EOCD3 )CBγ
S10 = − kDα A3αFCD3CF γ,
(25)
1 − Z15CF (10)
= kT1OCT1E CB , dCB3 dt = (Z10S7CF − kB2EOCB3 )CBγ
(26)
kT2CDC A2X (COH− CB ) + kT2EOCT3 + k2TCT2 − kBα A3αFCB3CF γ,
S11 =
1 − Z16CF + Z17 (11)
dCM3 dt = (Z13S9CF − kM2EOCM3 )CBγ
= kT2OCT2E CB , (27)
− kMα A3αFCM3CF γ,
where Z = k11 k1O , Z1 = k1E k1O , Z 2 = k2E k2O ,
Z 3 = k3E k3O , Z 4 = k1X k2O , Z 5 = k2X k3O , dCT3 dt = (Z16S11CF − kT2EOCT3 )CBγ
(28)
Z 6 = kD1E kD1O , Z 7 = kD2E kD2O , Z 8 = kD2O k2DO , − kTα A3αFCT3CF γ,
Z 9 = kB1E kB1O , Z10 = kB2E kB2O , Z11 = kB2O k2BO , dCPS dt = (S4 + S5 + S6 + S7 )CBγ
Z12 = kM1E kM1O , Z13 = kM2E kM2O ,
− (kD1CPC A1X + kD2CPC A2X + kB1CPFC A1X
Z14 = kM2O k2MO , Z15 = kT1E kT1O , Z16 = kT2E kT2O , (29)
+ kB2CPFC A2X )COH− γ + k41α A3αFC A3CF γ
and Z17 = kT2O k2TO are the ratios between the rate
+ k42α A3C A3γ,
2 2
constants of fast reactions.
dC A dt = [(S1 − ZC A ) − k1C A + k11OC AA dCDP dt = (S10 + S11)CBγ − (kT1CDPC A1X
(12) (30)
− kADC AC AA ]CBγ, + kT2CDPC A2X )COH− γ + kDα A3αFCD3CF γ,
dC AA dt = (ZS1C A − k11OC AA − kADC AC AA dCBP dt = kBα A3αFCB3CF γ, (31)
(13)
− 2kADC AA ) CBγ,
2

dCMP dt = kMα A3αFCM3CF γ, (32)

dC A1 dt = (Z1S1CF − k1EOC A1 − k2C A1 − S2 )CBγ, (14)
dCTP dt = kTα A3αFCT3CF γ, (33)
dC A2 dt = (Z 2S2CF − k2EOC A2 − k3C A2 − S3 )CBγ, (15)
dC A3 dt = (Z 3S3CF − k3EOC A3 − S3 )CBγ dCPN dt = k42α 2A3C A3
2
γ, (34)
(16)
− ( k41α A3αFC A3CF + 2k42α A3C A3 ) γ,
2 2
dCB dt = ( − k41C A3 − kDCD3 − kBCB3
(35)
dC AD dt = kADC AC AACBγ, − kMCM3 − kTCT3 − kFCF )α A3αFCF γ − k42α A3C A3γ,
2 2
(17)

KINETICS AND CATALYSIS Vol. 61 No. 3 2020

378
dCFS dt = (k1EOC A1 − Z1S1CF + k2EOC A2
− Z 2S2CF + k3EOC A3 − Z 3S3CF + kD1EOCD2
− Z 6S4CF + kD2EOCD3 − Z 9S6CF + kB1EOCB2
− Z 7S5CF + kB2EOCB3 − Z10S7CF kD2 +
+ kM1EOCM2 − Z12S8CF + kM2EOCM3 − Z13S9CF
+ kT1EOCT2 − Z15S10CF + kT2EOCT3
− Z15S11CF )CBγ − kB1CPFC A1XCOH− γ
− kB2CPFC A2XCOH− γ − (k41C A3 + kDCD3
+ 2kBCB3 + kMCM3 + kTCT3 )αFα A3CF γ
− k42α A3C A3CF γ − 2kFαFCF γ,
2 2 2 2
dCMS dt = kF αFCF γ + (S8 + S9 )CBγ −
2 2
− (kM1CMC A1X + kM2CMC A2X )COH− γ,
where CPS, CFS, and CMS are the total concentrations
of pentaerythritol, formaldehyde, and methanol, respec-
tively (mol/L), αF and αA3 are the degrees of ionization of
formaldehyde and 2,2-bis(hydroxymethyl)-3-hydroxy-
propanal, respectively, and γ is the activity coefficient
of singly charged ions.
The degrees of ionization of aldehydes were calcu-
lated by a method of successive approximations; the
concentration of hydroxyl ions was specified, and the
concentrations of aldehyde anions and methanol were
calculated:
C(i )COH− K a(i ) K W
C AN(i ) = , (38)
1 + COH− K a(i ) K W
where C(i) is the concentration of the ith substance
(mol/L), Ka(i) is the ionization constant of the ith sub-
stance, and KW is the ionic product of water. Then, the
concentration of hydroxyl anions corresponding to the
electroneutrality of solution was calculated as
N The parameters of the mathematical model were
COH−(1) = CB − C
i
AN(i ).
The calculation was carried out until the condition
COH− − COH−(1)
< X,
COH−
where X is the specified relative error of calculation
(X < 0.05).
The degree of ionization of aldehydes is
COH− K a(i ) K W
αi = . (41)
1 + COH− K a(i ) K W
According to published data [8],
pK w = 19.94 – 0.052T 0.9,
pK a(F) = 14.05 − 0.031T ,
pK a(A) = 14.05 − 0.037T ,
BELKIN, DEMCHENKO
pK a(AA) = pK a(A1) = 13.70 − 0.02T , (45)
pK a(A2) = 13.60 − 0.02T , (46)
pK a(A3) = 13.50 − 0.02T , (47)
pK a(M) = 16.30 − 0.31T , (48)
(36) where T is the temperature (°С).
The activity coefficient of singly charged ions was
calculated according to the Davis equation [9]
 
−logγi = Az 2  I − 0.2I  , (49)
1 + I 
where z is the ion charge (z = 1), A is a coefficient (A =
1.825 × 10–6[ε(T + 273)]–1.5), ε is the dielectric con-
(37) stant of the solvent (for water, ε = 78.55), I is the ionic
strength of the solution (I = CB(0)), and CB(0) is the ini-
tial concentration of sodium hydroxide (mol/L).
The rate constants k1, kAD, and kAD1 were previ-
ously estimated by studying the kinetics of acetalde-
hyde reactions in aqueous alkaline solutions at an ini-
tial acetaldehyde concentration of 0.25–1.2 mol/L, a
sodium hydroxide concentration of 0.01–0.1 mol/L,
and a temperature of 5–40°C. Sodium hydroxide is
consumed in the disproportionation reaction of
2-hydroxybutanal, whose rate constant is close to the
rate constant kF. This slow reaction was not taken into
account in the mathematical model.
Sodium hydroxide is not consumed in aldol con-
densation; this fact allowed us to separate the con-
sumption of aldehydes in aldol condensation and in
the Cannizzaro reaction:
C AL(A) = C AL + 2(CB(0) − CB ), (50)
where CAL is the total concentration of aldehydes
(mol/L).
(39) determined in the following sequence: First, condi-
tions under which the kinetics of the process mainly
depended on individual reactions were determined.
Under the selected conditions, we measured the cur-
rent concentrations of aldehydes, sodium hydroxide,
(40)
pentaerythritol, dipentaerythritol, bispentaerythritol-
formal, pentaerythritol methyl ether, and tripen-
taerythritol and the yields of products in three parallel
experiments. Then, we determined the arithmetic
mean values of the measured quantities and the mea-
surement error. Comparing the experimental and cal-
culated kinetic curves and the yields of products, we
preliminarily selected the rate constants of the test
reactions and determined the required rate constants
by the least-squares method.
(42) The rate constants of the slow stages of aldol con-
(43) densation k1, k2, and k3 were determined under the
conditions of minimal influence of side and reverse
(44) reactions at a molar ratio of 1 : 9–10 : 1.2 between
KINETICS AND CATALYSIS Vol. 61 No. 3 2020
 A MATHEMATICAL MODEL OF THE SYNTHESIS OF PENTAERYTHRITOL 379

(а) (b)
CAl, mol/L CAl, mol/L

4.5 CB, mol/L 4.5

1.0

4.0 4.0

1
3.5
3.5 1 CB, mol/L
0.5
0.5
3.0
3.0

2.5
2.5
0 0
0 50 100 0 100 200 300 400
t, min t, min

Fig. 1. Kinetics of synthesis with an excess of formaldehyde: (а) acetaldehyde–formaldehyde–sodium hydroxide molar ratio,
1 : 9 : 1.2; [acetaldehyde]0 = 0.60 mol/L; temperature, 20°C; constants (L mol–1 s–1) k1 = 0.031, k2 = 0.022, k3 = 0.018, and
k41 = 0.042; (b) acetaldehyde–formaldehyde–sodium hydroxide molar ratio, 1 : 8.5 : 1.04; [acetaldehyde]0 = 0.47 mol/L; tem-
perature, 6°С; constants (L mol–1 s–1) k1 = 0.039, k2 = 0.028, k3 = 0.023, and k41 = 0.063; (1) consumption of aldehydes in aldol
condensation, points and curves refer to experimental and calculation data, respectively.

acetaldehyde, formaldehyde, and sodium hydroxide, where k = 4.85 × 1015e–23.645/RT (L2 mol–2 h–1), and
an initial acetaldehyde concentration of 0.5–0.7 mol/L, 23.645 is the activation energy (kcal).
and a temperature of 5–40°C (Fig. 1). Our data show that the rate constant according to
The equilibrium constant of the aldol condensa- Eq. (52) remains relatively constant only at CB/CF =
tion of acetaldehyde is 0.2–0.3. At CB/CF = 0.05–2.0, the rate of reaction is
adequately described by the equation
C AA k11 k1 k
KA = = =Z 1 , (51)
CA
2
k1O k11O k11O dCB dt = −kF αF2 CF2 γ. (53)
where O refers to the reverse reaction. The value of the rate constant kF is given in Table 2,
According to Guthrie [10], KA = 400 L/mol and and Fig. 2 exemplifies a comparison between the cal-
k11О = 1.98 × 10–3 mol–1 s–1 at 25°C. Then, culated and experimental consumptions of sodium
−3
hydroxide and formaldehyde.
= 0.032 and Z = A 11O = 400 × 1.8 ×−10
k11O K k
= 13. Equation (53) corresponds to the generally
6.1 × 10
2
k1 k1 accepted Cannizzaro reaction mechanism, according
The constant kF was determined by studying the to which the formaldehyde anion and unhydrated
kinetics of formaldehyde reactions in aqueous alkaline formaldehyde, the concentration of which depends on
solutions at an initial formaldehyde concentration of the concentration of the anion, enter the reaction.
0.5–3.3 mol/L, an initial molar ratio of 0.2–1.5 At the found values of the constants kF, k1, k2, and
between sodium hydroxide and formaldehyde, and a k3, we determined the rate constants of the Cannizzaro
temperature of 20–50°C. reaction of 2,2-bis(hydroxymethyl)-3-hydroxypro-
According to Peters and Cupit [3], at an initial panal and formaldehyde and the disproportionation of
formaldehyde concentration of 3.3 mol/L and a sodium 2,2-bis(hydroxymethyl)-3-hydroxypropanal. With an
hydroxide concentration of 1 mol/L, the rate of the Can- excess of formaldehyde, the calculated and experi-
nizzaro reaction is described by the equation mental values of the current concentration of sodium
hydroxide coincided regardless of the rate constant k42
dCB dt = −kCBCF2 , (52) (Fig. 1); however, at a formaldehyde deficiency and

KINETICS AND CATALYSIS Vol. 61 No. 3 2020

380 BELKIN, DEMCHENKO

Table 2. Ratios between the rate constants of fast reactions

Parameter Value Parameter Value Parameter Value
Z 13 Z6 3.3 Z12 3.3
Z1 30 Z7 3.3 Z13 3.3
Z2 20 Z8 1.0 Z14 1.0
Z3 20 Z9 3.3 Z15 3.3
Z4 7.1 Z10 3.3 Z16 3.3
Z5 7.1 Z11 1.0 Z17 1.0

k42 = 0 (Fig. 3), a significant discrepancy was observed
to indicate a significant effect of the disproportion-
ation of 2,2-bis(hydroxymethyl)-3-hydroxypropanal
on the kinetics of the process.
The activation energy values found by us for the
slow stages of aldol condensation significantly differ
from those reported by Vik [11], and for the Canniz-
zaro reaction, they coincide with those given by Peters
and Cupit [3].
At a molar ratio of 1 : 3 : 0.9 between acetaldehyde,
formaldehyde, and sodium hydroxide, the specific
consumption of acetaldehyde in side reactions is
0.08–0.10 (mol/mol). With the found values of k1,
k11О, kAD, kAD1, and Z, the condition for this specific
consumption of acetaldehyde in side reactions is
k
Z1 = 1E ≈ 30.
k1O
An increase in the values of Z2 and Z3 above 15 does
not lead to a change in the calculated kinetics. Taking
into account the uniformity of the stages of aldol con-
k k
densation, we assumed Z2 = Z3 = 20 and 1EO = 2EO =
k1 k2
k3EO
= 0.05.
k3
The determination of the rate constants of reac-
tions in which dipentaerythritol, bispentaerythritol-
formal, pentaerythritol methyl ether, and tripen-
taerythritol are formed is a complex task. Therefore,
we selected the sets of effective rate constants (reflect-

(а) (b)
CF, mol/L

CF, mol/L
CB, mol/L 3 CB, mol/L
1.0 2 1.0 1.0

2
0.5 0.5 0.5

1
0 0
0 100 200 0 100 200
t, min t, min

Fig. 2. Kinetics of the formaldehyde disproportionation reaction: (a) 0–1 mol/L and (b) 1–3 mol/L. Temperature, 40°С: (1) calculation
according to Eq. (53) at kF = 3.2 × 10–4 L mol–1 s–1 and (2) calculation according to Eq. (52) at k = 2.33 × 10–4 L2 mol–2 s–1 [3],
points and curves refer to experimental and calculation data, respectively.

KINETICS AND CATALYSIS Vol. 61 No. 3 2020

 A MATHEMATICAL MODEL OF THE SYNTHESIS OF PENTAERYTHRITOL 381

CAl, mol/L CF
3
CTF = [12], (54)
2 2
0.026C W
K1
CFM = CFCM [13], (55)
K2
CB, mol/L CPS
CP = [2], (56)
1 0.5 1 + 0.05CF2.92
1
CPF = CPS − CP , (57)
CPF = CPS − CP , (58)
CM = CMS − CFM , (59)
0
0 100 200 where CF, CTF, CFM, CP, CPF, and CM are the equilib-
t, min rium concentrations of formaldehyde, its trimer, the
formaldehyde hemiacetal of methanol, pentaerythri-
tol, the formaldehyde hemiacetal of pentaerythritol,
Fig. 3. Kinetics of synthesis with a formaldehyde defi-
ciency: acetaldehyde–formaldehyde–sodium hydroxide
and methanol, respectively (mol/L); 0.026 is the equi-
molar ratio, 1 : 2.9 : 0.9; [acetaldehyde]0 = 0.5 mol/L; librium constant of the reaction of formaldehyde tri-
temperature, 16°C, constants (L mol–1 s–1) k1 = 0.017,
mer formation (almost independent of temperature in
k2 = 0.012, k3 = 0.01, k41 = 0.0025, and k42 = 0.0038; 2.9602 − 1798
a range of 0–50°C [12]); and K1 = 10 273+T and
(1) consumption of sodium hydroxide at k42 = 0, points
2806.45−9.947
and curves refer to experimental and calculation data,
respectively. K2 = 10 273+T are the equilibrium constants of
the reactions of unhydrated formaldehyde with water
and methanol, respectively (mol/L) [13].
ing the total effect of complex reactions) that ensured The equilibrium concentration of formaldehyde
good agreement between the calculated and experi- was calculated using the material balance equation
mental yields of pentaerythritol and the above prod-
ucts under various synthesis conditions, in particular, CF(1) = CFS − 3CTF − CFM − CPF . (60)
in the presence of added pentaerythritol. In this case, The calculation was carried out until the condition
the kinetic parameters of the same types of reaction
stages were taken equal: k41 = kD = kB = kM = kT, CF − CF(1)
< X, (61)
kD1 = kB1 = kT1, kD2 = kB2 = kT2, k2D = k2B = k2T, CF
kD1EO = kB1EO = kM1EO = kT1EO, kD2EO = kB2EO = where X is the specified relative error of calculation
kM2EO = kT2EO, Z4 = Z5, Z 6 = Z 7 = Z 9 = Z10 = (X < 0.05).
Z12 = Z13 = Z15 = Z16 , and Z 8 = Z11 = Z14 = Z17. The calculations did not take into account the
formaldehyde hemiacetals of dipentaerythritol due to
This solution seems to be quite correct because their low concentration.
simultaneous side reactions pass sequentially through Tables 2 and 3 summarize the ratios of the rate con-
the same stages of interactions between the same stants of fast stages and the kinetic parameters of the
groups of structurally similar reactant molecules and reactions, respectively.
their reactivity is not strongly different.
To evaluate the adequacy of the mathematical
The constants of hydration reactions of unsatu- model, we used a compositional plan with three fac-
rated aldehydes k1XO and k2XO were taken according to tors determining the results of the synthesis of pen-
Guthrie [10]. taerythritol: the molar ratio of formaldehyde and acet-
aldehyde M, the initial concentration of acetaldehyde
The equilibrium concentrations of formaldehyde, CA(0), and the temperature T (Table 1).
its trimer, methanol, the methanol hemiacetal of
formaldehyde, pentaerythritol, and the formaldehyde Table 4 summarizes the experimental and calcu-
hemiacetal of pentaerythritol in the mathematical lated values of the yields of pentaerythritol (P), dipen-
model were calculated by a successive approximation taerythritol (DP), bispentaerythritol (BP), pen-
method, when the equilibrium concentrations of taerythritol methyl ether (MP), and sodium 2,2-
formaldehyde were specified and the equilibrium con- bis(hydroxymethyl)-3-hydroxypropanoate (PN).
centrations of the other substances were calculated The variance of reproducibility was determined
from the formulas based on the results of six parallel experiments at the

KINETICS AND CATALYSIS Vol. 61 No. 3 2020

382 BELKIN, DEMCHENKO

Table 3. Kinetic parameters of pentaerythritol synthesis reactions

Rate constant Е, kJ/mol А, L mol–1 s–1 Rate constant Е, kJ/mol А, L mol–1 s–1
k1 99.8 1.88 × 1016 kD1 = kB1 = kT1 99.8 2.11 × 1016
k2 99.8 1.31 × 1016 kD2 = kB2 = kT2 99.8 1.74 × 1016
k3 99.8 1.09 × 1016 kM1 99.8 0.69 × 1016
k11O 99.8 0.60 × 1015 kM2 99.8 0.57 × 1016
k1EO 99.8 0.94 × 1015 k2D = k2B = k2M = k2T 99.8 1.09 × 1016
k2EO 99.8 0.65 × 1015 kD1EO = kB1EO = kM1EO = kT1EO 99.8 0.65 × 1015
k3EO 99.8 0.55 × 1015 kD2EO = kB2EO = kM2EO = kT2EO 99.8 0.55 × 1015
k41 = kD = kB = kM = kT 92.7 1.42 × 1014 kAD = kAD1 78.9 9.50 × 1012
k42 92.7 2.13 × 1014 kF 99.1 9.05 × 1012
k1XO 99.8 6.55 × 1016 k43 92.7 5.11 × 1013
k2XO 99.8 5.45 × 1016

Table 4. Plan and experimental and calculated results

Expe- Yield of product, %
riment M CA(0) T
number Pe Pc DPe DPc BPe BPc МРe МРc PNe PNc

1 + + + 79.1 77.9 2.9 3.1 3.1 2.7 4.1 3.6 10.8 12.3
2 – + + 71.2 72.9 6.2 5.7 0.5 0.8 2.7 2.9 19.4 17.7
3 + – + 82.0 80.8 3.0 3.4 1.8 1.4 3.1 2.9 10.1 11.5
4 – – + 73.1 74.2 6.1 5.2 0.2 0.2 2.2 2.4 18.4 18.0
5 + + – 79.1 77.0 2.2 2.5 2.7 2.4 3.0 3.2 13.0 14.9
6 – + – 70.8 71.6 6.2 5.5 0.6 0.8 3.0 2.7 19.4 19.4
7 + – – 80.2 79.3 2.3 2.6 1.5 1.2 2.8 2.5 13.2 14.4
8 – – – 72.5 73.0 5.4 5.4 0.1 0.2 2.6 2.3 19.4 19.6
9 +1.68 0 0 81.0 79.9 1.8 2.2 2.8 2.4 3.4 3.1 11.0 12.4
10 –1.68 0 0 67.3 68.0 7.5 6.9 0.0 0.1 2.2 2.5 23.0 22.5
11 0 +1.68 0 75.0 74.7 3.3 3.0 2.1 2.0 3.5 3.2 16.1 17.1
12 0 –1.68 0 77.8 78.2 3.6 3.7 0.0 0.4 2.0 2.3 16.6 15.4
13 0 0 +1.68 77.9 78.3 5.1 4.5 1.0 1.4 3.1 3.0 12.9 12.8
14 0 0 –1.68 74.2 75.6 3.0 3.6 1.5 1.3 2.1 2.5 19.2 17.0
15 0 0 0 76.1 77.1 4.1 3.8 0.8 1.2 2.1 2.7 16.9 15.2
16 0 0 0 75.7 77.1 3.7 3.8 1.4 1.2 2.9 2.7 16.3 15.2
17 0 0 0 78.0 77.1 3.4 3.8 1.1 1.2 2.4 2.7 15.1 15.2
18 0 0 0 78.2 77.1 3.9 3.8 1.3 1.2 2.9 2.7 13.7 15.2
19 0 0 0 75.2 77.1 3.2 3.8 1.5 1.2 2.3 2.7 17.8 15.2
20 0 0 0 75.8 77.1 3.6 3.8 0.9 1.2 2.0 2.7 17.7 15.2
Repeatability variance S y2 1.624 0.106 0.180 0.292 2.54

Residual sum of squares SR2 26.72 3.88 1.46 2.39 1.22

Variance of adequacy S A2 2.325 0.419 0.07 0.116 3.57

Calculated F value Fc 1.43 3.95 0.38 0.40 1.40
Tabulated F value FT 4.74 4.74 3.33 3.33 4.74
Designation of products is given in Scheme 1, and the subscripts e and c refer to experimental and calculated data, respectively.

KINETICS AND CATALYSIS Vol. 61 No. 3 2020

 A MATHEMATICAL MODEL OF THE SYNTHESIS OF PENTAERYTHRITOL
center of the plan. The tabulated Fisher’s values FT
were determined for a significance level of 0.05 at the
numbers of degrees of freedom related to the variance
of reproducibility
f y = N 0 − 1 = 6 − 1 = 5, (62)
where N0 is the number of experiments at the center of
the plan, and the variance of adequacy
(n + 2)(n + 1)
fA = N −
2 (63)
(3 + 2)(3 + 1)
= 20 − = 10,
2
where N is the number of experiments, and n is the
number of factors.
A comparison between the calculated and tabu-
lated Fisher test values indicates the adequacy of the
model.
It should be noted that a statistical assessment of
the adequacy of the model was performed on the basis
of linearization. With consideration for a large number
of parameters to be determined, their values may be
ambiguous.
CONCLUSIONS
We compiled a reaction scheme for the synthesis of
pentaerythritol by the condensation of formaldehyde and
acetaldehyde in the presence of sodium hydroxide.
We determined the kinetic parameters and activa-
tion energies of these reactions.
The compiled mathematical model of the synthesis
is adequate under the following conditions: an acetal-
dehyde–formaldehyde–sodium hydroxide molar
KINETICS AND CATALYSIS Vol. 61 No. 3 2020
383
ratio of 1 : 4–9 : 1.2, an initial acetaldehyde concentra-
tion of 0.4–0.9 mol/L, and a temperature of 13–47°C.
The mathematical model can be used to evaluate
the efficiency of synthesis in reactors of various types.
CONFLICT OF INTEREST
The authors have no conflicts of interest.
REFERENCES
1. Belkin, D.I., Zh. Prikl. Khim., 1990, vol. 63, no. 10,
p. 2312.
2. Demchenko, O.A. and Belkin, D.I., Visnik SNU
im. V. Dalya, 2011, no. 10, p. 42.
3. Peters, M.S. and Cupit, C.R., Chem. Eng. Sci., 1959,
vol. 10, p. 57.
4. Suchanec, R.R., Anal. Chem., 1965, vol. 37, p. 1361.
5. Trevoy, L.M. and Myers, M.E., Can. J. Chem., 1963,
vol. 41, p. 770.
6. Shilov, E.A., Zh. Prikl. Khim., 1935, vol. 8, p. 93.
7. Smith, B. and Tullberg, L., Acta Chem. Scand., 1965,
vol. 19, no. 3, p. 605.
8. Albert, A. and Serjeant, E.P., Ionization Constants of
Acids & Bases, New York: Wiley, 1962, p. 179.
9. Butler, J.N., Ionic Equilibrium: A Mathematical Approach,
Boston: Addison-Wesley, 1964.
10. Guthrie, J.P., Can. J. Chem., 1952, vol. 52, p. 2037.
11. Vik, J.E., Acta Chem. Scand., Ser. B, 1974, vol. 28,
p. 325.
12. Walker, J.F., Formaldehyde, New York: Reinhold, 1953,
p. 382.
13. Ogorodnikov, S.K., Formal’degid (Formaldehyde),
Leningrad: Khimiya, 1984, p. 89.
Translated by V. Makhlyarchuk